(E-journal) Plastic & Rubber Singapore Journal · Thiam Aik Tin-lead Alloy/Carbon Polymeric...
Transcript of (E-journal) Plastic & Rubber Singapore Journal · Thiam Aik Tin-lead Alloy/Carbon Polymeric...
-
1
[An official publication of the Plastic and Rubber Institute of Singapore]
Plastics & Rubber Singapore Journal [Volume 16]
The Plastic and Rubber Institute of Singapore
MITA(P)No:160/01/2016
-
2
-
3
CONTENTS Foreword
PRIS Management Committee 2015
Corporate members of PRIS
A Message from the President of PRIS
TECHNICAL PAPERS
A Practical Combination Test Method for Effectively Evaluation of the Rubber Materials for Fenders
By Kousik Kumar Mishra
Copolymerization Studied with MALDI ToF MS
By Alex van Herk
Electrode Nano-materials for Energy Storage and Conversion Application
By Tan Hui Teng, Alex Yan Qingyu, Ho Keen Hoe, Lee Hiang Wee, Ng Ming Xuan and Ho
Thiam Aik
Tin-lead Alloy/Carbon Polymeric Composites with High Electrical Conductivity
By Yongzheng Pan
Synthesis and Characterisation of Anode Nano-materials for Lithium Ion Batteries
By Tan Hui Teng, Alex Yan Qingyu, Kee Yong Yao, Chan Zing Weng, Kevin Leow Cheng Sheng
and Ho Thiam Aik
Film for Filtering Infra-Red Radiation from the Sunlight
By Kang Semi, Kim Jee Hyun, Murali Krishnaswamy and Hong Han
Volatile Corrosion Inhibitor (VAPPRO 872)
By Tong Shaw Wen, Geradine Yeo Li Yee, Quek Xin Lin, Ho Thiam Aik and Moe Patrick
Optimisation and Characterisation of Commercial Water-based Corrosion Inhibitor (VCI)
By H. Bryan Lim Han Yuan, Alex Low Shaw Boon, Cheang Tze Mun, Ho Thiam Aik and Moe
Patrick
-
4
Foreword
The Plastics and Rubber Institute of Singapore (PRIS) is a national academic organization
established for members who are in businesses related to plastic and rubbers. The members are
from manufacturing industries, trading companies, service business, education and research
institutes and students from polytechnics and universities. PRIS publishes journals periodically
aiming to meet the interest of and promote information exchange in the local industries and
within researchers pertaining to plastics, rubbers and additives as well as relate interesting high-
tech areas.
The committee of PRIS is publishing its 16th issue of Plastic and Rubber Singapore Journal.
This issue includes methods for characterization of rubber products and investigation of
copolymerization; nano-materials and polymer composites for energy applications, functional
films and inhibitors for corrosion control etc. I would like to take this opportunity to thank our
members, corporate members and all related persons and companies for their strong support. I
would like to thank all the authors for their dedicated contribution for this journal.
I am grateful to Dr. Zhao Jianhong and Dr. Ludger Paul Stubbs for reviewing the papers. I
am also grateful to Ms Jane Koh for her help and coordinate for the publication of the hard
copies soon. The support from Mr. Nee Pai How, Mr. Sanjeev Kumar, Dr. Gu Haiwen and Mr.
Ong Kian Soo is also appreciated.
Thank you.
Hong Han
Editor
-
5
PRIS Management Committee 2015
Dr Zhao Jianhong - President
Mr Ong Kian Soo - Vice-President / Social Secretary
Mr Sanjeev Kumar - Vice-President
Mr Nee Pai How - Honorary Secretary
Dr Gu Haiwen - Honorary Treasurer
Mr Ho Thiam Aik - Membership Secretary (Student Affairs)
Dr Hong Han - Publication Secretary
Dr Ludger Paul Stubbs - Technical Secretary
Dr Pan Yongzheng - Education Secretary
Dr Daniel Wang SF - Committee Member
Dr Lau Soo Khim - Committee Member
Mr Shi Junhao - Committee Member
Corporate Members and Supporting Companies and organizations
Apollo International Limited Chemart (S) Pte Ltd HLN Rubber Products Pte Ltd LMS Technologies Pte Ltd Purac Asia Pacific Pte Ltd Quantum Technologies Global Pte Ltd RTP Company (S) Pte Ltd Trelleborg Marine Systems Asia Pte Ltd Mindtrac (S) Pte Ltd East Chemical Pte Ltd Maha Chemicals (Asia) Pte Ltd Zwick Asia Pte Ltd TUV SUD PSB Chatsworth Associates Pte Ltd Singapore Polytechnic Institute of material Research and Engineering Singapore (IMRE) Institute of Chemical & Engineering Sciences Singapore (ICES) Singapore Institute of Manufacturing Technology (SIMTech) Oil & Colour Chemists' Association (OCCA) S'pore Chemical Industry Council Limited (SCIC) Plastics & Rubber Institute Malaysia (PRIM)
-
6
A Message from the President of PRIS
The Plastics and Rubber Institute of Singapore (PRIS) is very proud to present you this issue of Plastics
and Rubber Singapore – an official technical journal of the Institute.
Plastics and rubbers are nowadays widely used in all walks of our life playing irreplaceable roles. Though
some of the manufacturing related processes have moved out of Singapore, polymeric material
applications, the science and technologies are getting more and more important for our future
development. As the national professional society for the plastic and rubber industry, PRIS has
consistently devoted over the past 36 years since its establishment to the cause of developing,
promoting, and introducing new and advanced polymer science and technologies to the Singapore
plastics and rubber industry.
PRIS acts as the centre for promoting the interests of its members through a variety of technical
activities such as regular workshop, seminars, conferences, training courses, and visits to manufacturing
plants and industrial sites. We also organize various networking and entertaining events for members to
communicate and exchange view in friendly and relaxed environment. Publication of the Journal has
served as a means for technical communications, exchange of views, and promotion of new
technologies. This issue of the Journal has included a few selected papers that were presented at our
meeting and seminars in the past year for the benefit of our members and readers; it is also glad to see
that a few papers are authored by teachers and students from Singapore Polytechnic.
I would like to take this opportunity to thank all PRIS members for their support rendered to the
Management Committee over the past years. I would also like to congratulate Dr. Hong Han, the Editor
of the Journal, and the Journal sub-Committee members Dr. Ludger Stubs, Dr. Pan Yongzheng, Mr. Ong
Kian Soo, and Ms Jane Koh, for the excellent job done throughout the process of the publication.
Finally, on behalf of the Institute, I would like to thank our corporate members for their supports in the
publication of the journal.
Dr. Zhao Jianhong
President
The Plastics and Rubber Institute of Singapore
-
7
A Practical Combination Test Method for Effectively Evaluation of the
Rubber Materials for Fenders
Kousik Kumar Mishra
4 Jalan Pesawat Singapore 619362
Abstract A reliable, viable analytical method to assess the composition of recently procured fenders prior to delivery has been developed, using a simple sampling procedure from the surface of the fender. This new technique will help to ensure that fenders supplied use the correct quality of rubber compound required to adhere to the specification. The recommended tests to evaluate the quality of fenders, based on a sample of
only 20-50grams which can be easily gathered by obtaining scrapings from the final product prior to installation, without affecting the fenders performance during application.
Keywords: fender; quality; recycled rubber; test.
1. Introduction
1.1 Fender system:
Fendering systems are mission critical equipment for marine environments globally. Long gone are the days of
wooden or rope fenders, and the use of rubber has become standard best practice. Although rubber fenders have a
long lifecycle, ultimately it is still limited. Depending on the environment and quality of the fender itself, the
expected average lifespan will be approximately 15 to 20 years. Designing a fender system requires engineers to
determine the berthing energy of a vessel or range of vessels that are likely to be docked against the system, then
determining what capacity the fender needs to have to absorb that kinetic energy. Finally, engineers must find a way
to avoid creating too much force and damaging either the wharf structure or vessel.
It’s accepted that high quality fenders can add value to port operations as sourcing quality materials and fully
tested compounds allows ports to drive cost efficiencies, minimise maintenance requirements and reducing the risk of incidents. High quality fenders also have a longer service life and, due to reduced maintenance requirements, also
lead to fewer “lost” days for ports, and their shipping operators. In addition to these commercial concerns, fendering
systems are a port’s first line of defence when a vessel comes into dock and play a key role in protecting the safety
of port personnel, vessel crew, cargo and infrastructure.
However, there has been a worrying trend becoming more pronounced across the industry in recent years, of
putting up front costs higher on the agenda than whole life costs. Although this enables immediate cost savings for
procurement managers it means that, over the course of the fender’s lifecycle, costs will be higher. Some
unscrupulous fender suppliers are taking the opportunity to undercut reputable fender manufacturers by supplying
lower cost, but lower quality fenders. They are able to elicit cost savings to pass on to their customers (in the
immediate term) in two ways:
By using a higher percentage of recycled rubber within the fenders, instead of virgin rubber
Replacing carbon black fillers with non-reinforcing fillers A simplified comparison chart representing the whole life cost differences between the two can be found
below:
Spending requirements Low quality fender High quality fender
Purchase price of a CONE 1000 fender $8,000 $10000
Installation $4000 $4000
Replacement after five years $9200 n/a
Re-installation $6000 n/a
Maintenance $9000 $4000
Maintenance installation $12000 $4800
Total 10 year whole life cost $48200 $22800
-
8
Spend
These manufacturers also ‘copy’ correction factors from reputable manufacturers without understanding these
factors are derived based on both rubber compound used as well as fender geometry. They do not make any
investment in PIANC Type approved fenders.
Low quality fender High quality fender
1.2 Setting the Standards “Although PIANC is not in a position where it may “regulate the industry” or deliver any certification, PIANC
is very careful to promote best practice. We also stress that in the long term, through life cycle approaches, it is
recommended to use the most adapted, strong and resistant fender protection to quays” mentioned by Geoffrey
Caude, PIANC President, 2011. There are a number of different standards used worldwide to design fender systems
but the most commonly used is PIANC’s “Guidelines for the design of fender systems, 2002”, which was updated
from its predecessor of 1984. Although PIANC set out these guidelines, they do not regulate the industry, or indeed,
enforce the guidelines in practical terms. This has led to some fender suppliers misusing PIANC “certification” by
applying it to fenders that use higher percentages of filler and recycled rubber than is appropriate. PIANC’s
guidelines specify that robust material testing is a necessity, and the fact that this is not routinely performed by all
suppliers as part of their quality assurance process is a serious concern. Laboratory and full scale testing are
fundamental to the design and production of mission critical equipment and the industry needs the reassurance that
both sets of testing have been performed. Some suppliers are able to cut costs though replacing natural rubber with reclaimed rubber, and using large amounts of non-reinforcing fillers, which is a poor substitute for the carbon black
reinforcing filler used in high quality fenders. These lower cost fenders, therefore, do not meet the required
specifications, won’t perform adequately whilst they’re in use and, as such, won’t have the product lifecycle they are
claimed to have. Additionally, port owners, contractors and consultants have no simple method available to test the
quality of the fender’s material once it is purchased and installed.
A new analytical test has been developed to help buyers understand and substantiate what is in a fender and
ensure that port owners, operators and contractors can ensure the highest quality of fenders going forward. Both
chemical and physical testing are required to verify the rubber quality of the fender and ensure that it remains stable
and suitable for the use it was intended for, throughout its lifecycle, to ensure maximum protection of the port
infrastructure and the vessels that come to berth there.
2. Methods
To demonstrate and quantify the difference in performance characteristics of a high quality and low quality fender, the following tests were carried out in an independent third party laboratory:
Comparison of the physical properties of the rubber samples. The samples were cut from two commercial sized fenders: one a typical high quality fender, and one a typical low cost fender
Comparison of the chemical properties of the fenders. The samples were taken from the fender surface of two commercial sized fenders: one a typical high quality fender, and one a typical low cost fender.
The following tests were conducted:
Physical analysis:Test Equipment Used Expected Standard
Density Weighing balance ISO 2781
Hardness Shore A hardness tester ASTM D2240
Tensile strength Universal test machine ASTM D412
Elongation at break Universal test machine ASTM D412
Chemical analysis:
Test Equipment Used Expected Standard
Polymer (virgin plus recycled rubber) % TGA /FTIR ASTM D6370/D297
Carbon black % TGA /FTIR ASTM D6370/D297
Ash % TGA ASTM D297
Calcium Carbonate (white filler %) Chemical method ASTM D297
1. For further information on TGA and FTIR equipment, please see footnote [2].
-
9
3. Results and discussion
Physical analysis:
Test Standard High quality
fender
Low cost
fender
Requirement
Density (g/cc) ISO 2781 1.15 1.29 Not specified
Hardness (shore A) ASTM D2240 67 71 Max. 78
Tensile Strength (Mpa) ASTM D2240 15.4 9.3 15.2 - 13.6 (Note)
Elongation @ Break (%) ASTM D2240 364 278 297 - 333 (Note)
2. For more detailed information on results, see footnote [3].
The cost of a fender is often reduced by using a higher percentage of recycled rubber, and low cost non-
reinforcing white calcium carbonate (CaCO3) fillers in the formulation. We found that fenders with recycled rubber
and filler are heavier (and denser) than virgin rubber fenders. This significant weight difference enables a user to
evaluate whether a fender uses low cost recycled materials, or is the genuine article, made with high performance
rubber compound, with the benefits of long life and superior resilience. Chemical and physical analyses revealed
some further interesting insights into the materials used for manufacturing the fenders, and the properties these
materials have:
Values of tensile strength and elongation at break for the low cost fender were lower than the high quality fender and not in compliance with the user specification.
Rubber to filler ratio (Polymer %: Carbon Black % + Ash %) for the high quality fender was 1.23. This simply means 1 kg of filler was blended with 1.23 kg of rubber. The rubber to filler ratio for low cost
fender was only 0.88, which means 1 kg of filler was blended with just 0.88 kg of rubber.
The low cost fender contained 28.45% less rubber than high quality fenders. The presence of more rubber in high quality fenders explains the reason behind better physical properties of these fenders, and also
justifies the higher cost. For perspective, the cost of rubber is usually three times higher than fillers like
carbon black.
-
10
Most of the raw rubbers are weak when vulcanized and need reinforcing filler to increase mechanical properties of the final product. Ash analysis of the high quality fender indicated that it contains 100%
carbon black filler which is high quality reinforcing filler.
On the other hand, the ash analysis result of the low cost fender showed presence of only 55% carbon black and 45% CaCO3. The price of CaCO3 is approximately a fifth cheaper than carbon black. CaCO3 is
considered as white, non-reinforcing filler which is usually used to reduce the cost of the rubber compound but does not help in improving the properties.
The density of low cost fender is 12% higher than the high quality fender. The reason behind the higher density of the lost cost fender could be attributed to the following two factors:
– The presence of high density CaCO3 in the formula at 15.54%, as determined by ash analysis. Note that
density of CaCO3 is 2.7 g/cc while that of Carbon black is 1.8g/cc.
– The presence of a high percentage of recycled rubber in the formulation is the other contributing factor.
The density of recycled rubber is 1.15 to 1.20 g/cc while that of virgin rubber is 0.92 g/cc.
Recycling of rubber is a hard line, energy intensive process in which rubber powder is cooked with aggressive
chemicals. This process breaks long rubber molecules into shorter ones and thereby reduces the physical properties.
Usually tensile strength of recycled rubber is one-third of virgin Natural rubber (NR). Chemical analysis showed that the low cost fender contained 60% NR. However, the low tensile strength, elongation at break and high density
of the fender pointed towards the presence of high percentage of recycled rubber instead of virgin rubber.
4. conclusion
These newly developed physical and chemical tests provide a reliable, viable analytical method which can now
be made available for buyers to be able to assess the composition of recently procured fenders prior to delivery,
using a simple sampling procedure from the surface of the fender. This new technique will help to ensure that
fenders supplied use the correct quality of rubber compound required to adhere to the specification. The
recommended tests to evaluate the quality of fenders, based on a sample of only 20-50grams are listed in the table
below. These samples can be easily gathered by obtaining scrapings from the final product prior to installation,
without affecting the fenders performance during application.
Test Standard Specification
Test Standard Specification
Density ISO 2781 Max 1.20 g/cc
Polymer % ASTM D6370 Min. 45%
Carbon Black % ASTM D6370 Min 30%
Ash % ASTM D297 Max 5%
Chemical testing is not enough the guarantee fender performance and full scale testing should also be
performed in the factory to guarantee the lifecycle and performance of fenders meet the specification they are
intended for. As demonstrated, manufacturers with in house design and engineering capabilities are able to test their
compounds in the laboratory and provide full scale testing on prototypes and finished products. It’s therefore imperative that port owners and specifies understand the importance of not making procurement decisions purely
based on up-front costs.
The equipment will need to be replaced earlier, and in the long term, require heavier investment, not to
mention the higher risks of failure during service life. Decision makers should be aware of these key differences and
the varying quality on offer when buying on the basis of short term cost savings. There is a need for the whole
industry to come together to discuss changes to a culture that is causing unprecedented levels of downtime and
putting ports at risk.
5. Reference and footnotes
[1]. Analytical equipment like TGA/FTIR are not usually used in testing for the fender industry for quality control checks, but were applicable in this case to enable chemical analysis of the rubber compounds.
-
11
TGA: Thermogravimetric Analysis measures the amount and rate of change in the weight of a material as a function of temperature or time in a controlled environment. Measurements are used primarily to determine the composition and predict thermal stability at temperatures up to 1000°C. The technique can characterize substances that exhibit weight loss or gain due to decomposition, oxidation or dehydration. FTIR: Fourier Transform Infrared Spectroscopy is most useful for identifying chemicals that are either organic or inorganic
in nature. It can be utilized to quantify some components of an unknown mixture. It can be applied to the analysis of solids, liquids and gases. The term Fourier Transform Infrared Spectroscopy refers to a fairly recent development of the manner in which the data is collected and converted from an interference pattern to a spectrum. Today’s FTIR instruments are computerized which makes them faster and more sensitive and accurate for composition analysis. NB: TGA/FTIR are unable to differentiate between virgin and recycled rubber generated from natural rubber.
[2]. It is assumed that tensile strength and elongation at break of samples prepared from cured product will be 5-15% lower than samples prepared in the laboratory by moulding of uncured rubber
[3]. Specification: Tensile Strength Min 16Mpa, Elongation at break 350% min. (Ref: Physical Testing of Rubber by Roger
Brown, Chapter 3, page 47). The values reported were median of five reading. Tensile strength 16mpa min, E@B 350% min when tested in the compound. These values are lower when tested in the sample taken from final produce (fender). 20192
_______________________ Mr Kousik Kumar Mishra is currently working with Trelleborg, a Sweden based company and market leader in non-tyre industrial products, since 2001. Mishra was responsible for the Technical development of Marine fenders, Mining products and general purpose & special purpose industrial goods for Trelleborg Singapore, China & Australia manufacturing units. Currently He is working as Global Technical and Market Support Manager for Trelleborg Marine Systems and supporting technical and sales team globally in fender business. With more than 15 years of experience in the rubber industry, Mishra has presented technical papers in various international conferences. Mishra holds a B.Tech in Rubber Technology from University of Calcutta, M.Sc (Tech) from UDCT, Mumbai University and an Executive MBA from the Chicago Booth School of Science.
-
12
Copolymerizations Studied with MALDI ToF MS
Alex van Herk
Institute of Chemical and Engineering Sciences 1 Pesek Road, Jurong Island Singapore, 627833
Email: [email protected]
Abstract A reliable method based on a single MALDI-ToF-MS spectrum of the copolymers was applied to determine reactivity ratios. Since MALDI-ToF-MS gives information on individual polymer chains, access to homo-propagation and cross-propagation probabilities becomes available. These probabilities provide the reactivity ratios by simulation of a first order Markov chain by using the Monte Carlo method. The
experimental results of random copolymers of P(styrene-co-butylacrylate) was reported. The reactivity ratio of styrene and butylacrylate was calculated as 0.79 and 0.21 respectively.
Keywords: Copolymers; MALDI-ToF-MS spectrum; reactivity ratio.
1. Introduction
One of the ultimate challenges in polymer chemistry is the ability to control the physical properties of a
copolymer by tailoring its microstructure. Knowing the reactivity ratio of the comonomers allows predicting and
tuning of the copolymer’s microstructure, both with respect to composition and topology. The classical method to
ascertain reactivity ratios is by determining the comonomer composition of a range of polymers prepared with
different feed compositions. A fast and reliable method that can prevent this tedious and time-consuming laboratory
work is therefore highly desired. Various methods to determine reactivity ratios have been reported which deal with either the differential or the integral form of the Mayo-Lewis equation [1]. Nevertheless, most methods have the
disadvantage that still quite some reactions have to be performed with different feed compositions. Moreover,
comparison of ratios obtained by diverse methods often shows a relatively big variety due to a statistical error by for
example linearization of the equations and absence of weighing the individual data. Choosing the right statistical
method is therefore crucial for the reliability of the outcome [2].
Since a copolymer is a statistical mixture of individual molecules, a copolymer sample obtained from a single
experiment in principle contains all the information required to retrieve the reactivity ratios. Still, examples of
methods that only require a single experiment are limited. Jaacks introduced a method in which the ratios are
determined from a single experiment when one of the two monomers is in large excess [3]. This method is limited to
systems in which the reactivity ratios do not have an extreme difference in values [4]. Rudin reported on the use of a
single NMR spectrum by using the sequence distribution as determined from the measured diads or triads [5]. However, to obtain a highly resolved spectrum, long measuring times are required and assigning the peaks is not
straightforward. In ICES we recently acquired a MALDI-ToF MS and have implemented a new method to obtain
information about copolymerizations. Matrix assisted laser desorption/ionization time-of-flight mass spectrometry
(MALDI-ToF-MS) is a fast and accurate technique to determine copolymer compositions and in principle should be
suitable to rapidly determine reactivity ratios of comonomers. The principle of MALDI ToF MS is that polymer
chains are brought into flight in a mass spectrometer by using a matrix material for the sample that easily evaporates
when illuminated by a laser beam and with it, it brings the polymer chains into flight. A salt is added to bring the
polymer a charge.
Recently, we have reported on the use of MALDI-ToF-MS to determine polymer topologies and to study
mechanistic aspects of various copolymerization systems [6,7]. The first to apply MALDI-ToF-MS to determine reactivity ratios of comonomers were Suddaby and Willemse but they still required data from different reactions
[8,9].
mailto:[email protected]
-
13
In this study we are going to apply a reliable method to determine reactivity ratios based on a single MALDI-
ToF-MS spectrum of the copolymers. Since MALDI-ToF-MS gives information on individual polymer chains,
access to homo-propagation and cross-propagation probabilities becomes available [10]. These probabilities provide
the reactivity ratios by simulation of a first order Markov chain by using the Monte Carlo method. The reactivity
ratios have been determined with this new method successfully for three different types of copolymerizations i.e.
free radical polymerization, ring-opening polymerization of cyclic esters and of oxiranes and anhydrides [11].
2. Methodology and Results
MALDI-ToF-MS spectra give highly accurate molar masses of all the polymer chains in the sample which not
only enables elucidation of individual chain lengths, but provides full characterization including the copolymer’s
chemical composition and to some extend the copolymer topology (random, gradient, block, alternating). MALDI-
ToF-MS spectra can be deconvoluted by employing the equation:
MEEMwmMwnm IIIcal 2211 (1)
where EI and EII represent the molar masses of the end groups at opposite sides of the chain, n1Mw1 and m2Mw2
represent the number and molar mass of the repeating units of monomer M1 and M2 respectively, and M+ the mass
of the cation (a salt is usually added for charging the polymer chains in order for them to be accelerated in the mass
spectrometer). With this equation, a complete matrix with n1,i rows and m2,j columns can be constructed for a given end group combination. (see Figure 1). The peaks in the spectrum are assigned to a certain position in the matrix
employing the inequality:
(2)
In which mexp represents the experimental mass, mcal the calculated mass and Δm the accuracy (1-2 g·mol1). By
calculating the natural abundance isotope distributions for each position in the matrix and rescaling it to the
corresponding highest-intensity mass-peak, a full spectrum can be simulated as well as the corresponding contour
plot, which provides information about the polymer composition (see Figure 2).
Figure 1. Schematic representation of the matrix of the copolymer.
The chemical composition distributions for the chain lengths covered by the total chain length distribution can
be obtained by diagonally walking through the matrix from (0,M1,n) to (M2,n,0) after normalization using the sum of
intensities within this chain length. The distribution of monomer repeating units along an individual chain can be
2exp
mmm cal
-
14
described by a first order Markov chain by means of the Mayo-Lewis (terminal) model [11]. In this study we use the
terminal model for copolymers in which we can distinguish the following four probabilities:
221
21
21
221
2
22
211
12
211
211
11
r]M/[]M[
]M/[]M[P
r]M/[]M[
rP
1])M/[]M([r
1P
1])M/[]M([r
])M/[]M([rP
21
*
121
*
2
22
*
222
*
2
12
*
212
*
1
11
*
111
*
1
kMM~MM~
kMM~MM~
kMM~MM~
kMM~MM~
The reactivity ratios are then given by r1= P11/P12 [M2]/[M1] and r2= P22/P21 [M1]/[M2].
The reactivity ratios can be obtained from the MALDI-ToF MS spectrum at one chain length at a time (see slice in
Figure 2), also allowing to look for chain length dependent effects, still from one single experiment.
Figure 2: Copolymer matrix for a styrene-butylacrylate copolymer represented in a contour plot and one chemical
composition distribution (CCD) at chain length 20 monomer units (right) also showing the fitted CCD resulting in the reactivity ratios for this copolymerization system.
3. Reference [1] a) C. Hagiopol, Copolymerization, Towards a systematic approach, Springer, 1999. b) A.L. Polic, T.A. Duever, A. Penlidis,
J. Polym. Sci., Part A: Polym. Chem. 1998, 36, 813. c) M. Finemann, S.D. Ross, J. Polym. Sci., 1950, 5, 259. d) T. Kelen, F. Tüdös, J. Macromol. Sci., Chemistry, 1975, A9, 1.
[2] a) R. van der Meer, H.N. Linssen, A.L., German, J. Polym. Sci., Part A: Polym. Chem. 1978, 16, 2915. b) A.M. van Herk, T. Dröge (1997) Macromol. Theory Simul. 6 1263-1276. c) A.M. van Herk (1995) J. Chem. Ed. 72 138-140
[3] V. Jaacks, Makromol. Chem. 1972, 161, 161.
[4] D. R. Burfield, C. M. Savariar, J. Polym. Sci., Polym. Lett. Ed. 1982, 20, 515. [5] A. Rudin, K. F. O' Driscoll, M. S. Rumack, Polymer 1981, 22, 740. [6] S. Huijser, B.B.P. Staal, J. Huang, R. Duchteau, C.E. Koning, Angew. Chem. Int. Ed. 2006, 45(25), 4104. [7] S. Huijser, B.B.P. Staal, J. Huang, R. Duchteau, C.E. Koning, Biomacromolecules 2006, 7(9), 2465. [8] K. G. Suddaby, K. H. Hunt, D. M. Haddleton, Macromolecules 1996, 29, 8642. [9] R. X. E. Willemse, A. M. Van Herk, J. Am. Chem. Soc. 2006, 128, 4471. [10] M. S. Montaudo, A. Ballistreri, G. Montaudo, Macromolecules 1991, 24, 5051. [11] S. Huijser, G.D. Mooiweer, R. van der Hofstad, B.B.P. Staal, J. Feenstra, A.M. van Herk, C.E. Koning, R. Duchateau
(2012) Macromolecules 45, 4500-4510
Alex van Herk (1956) is senior researcher at the Institute of Chemical and Engineering Sciences in Singapore since
2012 and part-time professor in Polymer Reaction Engineering at the Eindhoven University of Technology, the
Netherlands (where he worked full-time from 1986 till 2012). Since 2009 he also is teaching at NUS regularly. His
field of research is nanotechnology, water-based coatings and emulsion polymerization. He is editor of four books
and author of 180 papers. At present he is chairman of the Foundation Emulsion Polymerization, a multisponsored
liaison program between industry and academia.
rST
=0.79
rBA
=0.21
-
15
Electrode Nanomaterials for Energy Storage and Conversion Applications
H. T. Tan1, Q. Y. Yan Alex1, K. H. Ho2, H. W. Lee2, M. X. Ng2, T. A. Ho2
1School of Material Science and Engineering Nanyang Technological University
2Diploma in Materials Science
School of Chemical & Life Sciences Singapore Polytechnic
Abstract Manganese oxide nanoparticles, a pseudocapacitor material, can be synthesised through hydrothermal synthesis to control and produce nanoparticles of desired morphologies. The addition of carbon nanotubes, a good electric conductor, into this synthesis process allows these nanoparticles to be form on these nanotubes and be embedded within them. This overcomes the poor electrical conductivity limitation of
manganese oxide, thus allowing its excellent pseudocapacitance properties to be harnessed. In hope to optimise the pseudocapacitance of this nanocomposite, the nanoparticle crystal structure, morphology and the effects of nanoparticle’s morphology on specific capacitance under varying parameters of synthesis time, temperature as well as the volume of potassium permanganate were respectively characterised and studied using X-ray diffraction, scanning electron microscopy and cyclic voltammetry. Post studies shows the synthesis of nanoparticles with flower, hexagonal plates and nanorods morphologies where flowers possess birnessite crystal structure while hexagonal plates and nanorods possess nsutite crystal structure. Optimal specific capacitance was shown to be achieved at 1600C, 120 minutes and using 2mL of 0.25M potassium permanganate where the highest specific capacitance obtained was 335F/g.
Keywords: Manganese oxide nanoparticles; carbon nanotubes; pseudocapacitance; morphology; specific capacitance.
1. Introduction
Environmental issues and depleting fossil fuels have sparked interest in the research and development of
alternative energy storage/conversion devices in recent years. Supercapacitors, or electrochemical capacitors, have
received enormous attention owing to their potential applications ranging from mobile devices to electric vehicles.
Supercapacitors are broadly classified into double-layer capacitors and pseudocapacitors with each having different
mechanisms of energy storage. Double-layer capacitors store energy via non-Faradic accumulation of charges at the
electrolyte-electrode interface while pseudocapacitors store charges by undergoing Faradic reactions which involve
the transfer of electrons [4]. Pseudocapacitors display higher specific capacitance over double-layer capacitors and
thus, various transition metal oxides have been investigated as potential pseudocapacitor materials [3].
Out of the various oxides investigated, hydrated ruthenium oxide (RuO2.nH2O) was reported to have specific
capacitances of up to 700F/g [4]. However, the use of ruthenium oxide is limited by its high cost, toxicity, and the required use of highly acidic electrolytes to obtain peak performance. On the other hand, manganese oxide (MnO2)
nanoparticles (NPs) which have the advantages of low synthesis cost, abundance, non-toxicity and ability to perform
well in neutral electrolyte systems, has recently garnered attention as a promising pseudocapacitor material.
Although hydrated MnO2 has a theoretical specific capacitance value of 1370 F/g, it has been experimentally
reported to exhibit specific capacitances within 100 – 200 F/g which is far from the theoretical value of 1370 F/g
owing to its poor electrical conductivity [6].
MnO2 can be synthesised using different techniques, such as simple reduction, sol-gel, co-precipitation and in
particular, hydrothermal synthesis. Hydrothermal synthesis was found to be a good technique in the preparation of
nanomaterials with different morphologies such as wires, rods, urchins and belts. The main advantages of
hydrothermal synthesis over the other synthesis routes are its ability to have good control over the morphologies of
the nanoparticles formed as well as environmental benign since water is used as the solvent. Each nanostructure has its own advantages when used in potential applications. In the synthesis of MnO2 NPs, it was observed that the NPs
could take up 3 different morphologies which are highly relevant to the discussion of this study [6]. The NPs would
initially aggregate to form spherical flower agglomerates which with prolonged durations and specific conditions of
ageing, would gradually transform into nanorods. This transformation was due to the fact that the high specific
surface areas of nanospheres/nanoflowers led to high surface energies, and thus they would aggregate further to
-
16
form structures such as nanorods as they are more stable and have lower surface energies. This transformation is
consistent with the Ostwald Ripening process which stated that larger particles with lesser surface energy form at the
expense of smaller particles with greater surface energy. The third morphology observed was hexagonal plates
morphology and this morphology was regarded as an intermediate of the transformation from nanoflowers to
nanorods. Hexagonal plates appeared due to insufficient time and temperature for the transformation to occur
completely. Thus, it could be deduced that nanoflowers have the highest surface energies, followed by the hexagonal plates, and then nanorods. Generally, nanoflowers are desired as it has higher surface area available for surface
faradic reaction than hexagonal plates and nanorods morphology which therefore allow higher pseudocapacitance to
be obtained.
There has been an increasing use of carbon additives, such as carbon nanofoams, graphenes and CNTs, during
the synthesis of MnO2 NPs to overcome its poor electrical conductivity and improve its performance as a
supercapacitor. Due to their high electrical conductivity, such carbon additives are added to facilitate electron
transport to the MnO2 NPs during Faradic reactions.
This study aims to deduce the optimum synthesis of MnO2-CNT composite with the highest specific
capacitance. MnO2 was synthesised onto functionalised multi-wall CNTs through hydrothermal synthesis with
different reaction conditions, namely synthesis time, synthesis temperature, and the volume of potassium
permanganate (KMnO4) used for the synthesis. The effects of the reaction parameters on the morphology, the crystal
structure of the morphology produced and morphology on specific capacitance of the MnO2-CNT composite are studied and discussed.
2. Experimental
2.1 Preparation of Hydrothermal Synthesis
Figure 2.1.1: Hydrothermal synthesis procedure
Table 2.1.2: Samples and their respective parameters
Parameter 3: Time (Min)
45 120 180
Parameter
1: Volume
of 0.25M
KMnO4
(mL)
0.4
Parameter 2:
Temperature (°C)
120 A1 A2 A3
160 A4 A5 A6
200 A7 A8 A9
1
Parameter 2:
Temperature (°C)
120 B1 B2 B3
160 B4 B5 B6
200 B7 B8 B9
2
Parameter 2:
Temperature (°C)
120 C1 C2 C3
160 C4 C5 C6
200 C7 C8 C9
50mg of FCNT and 5mL of 1% w/v aqeous manganese (II) nitrate tetrahydrate were added into an autoclave's
Teflon capsule. After which the desired volume of the variable 0.25M KMnO4 were added before adding ultra pure
water till the solution reaches 50ml in total. The resulting solution is then sonicated for 5mins. The autoclave with its
1) 50mg FCNT
2) 33mL ultra-pure water
3) 5mL Manganese (II) Nitrate Tetrahydrate
4) Potassium Permanganate
Autocla
ve
Oven Centrifug
e
Removal of
supernatant
Repeat twice
-
17
Teflon capsule inside was subsequently placed and heated in an oven at the desired temperature and duration. Next,
the synthesised mixture was transferred into a centrifuge tube and subjected to centrifugation. After the first cycle
ends, the supernatant was removed and the centrifuge tube was topped up with ultra pure water till the solution
reaches 50ml before the next centrifugation. Another 2 cycles of centrifugation, refilling of ultra-pure water and
removal of supernatant were carried out in order to remove reactants from the sample. This procedure as illustrated
in figure 2.1.1 is repeated for the all the 36 samples with their respectively parameter as listed below in table 2.1.2.
2.2 Scanning Electron Microscope (SEM) analysis
The centrifuge tube containing our sample was topped up with ultra-pure water before ultra-sonicating the
solution. A micropipette was used to draw small amount of sample from the test tube and dropped on the shiny
surface of silicon wafer. The silicon wafer was dried in an oven at 70°C before inserting into JEOL JSM-7600F FE-
SEM for viewing of the sample's NPs' morphology. This procedure was repeated for each sample.
2.3 Cyclic Voltammetry (CV) test A mortar was used to grind 12mg of sample and 1.5mg of carbon black for 10 minutes. A slurry was obtained
by adding 3 drops of NMP and 25mg of 5% w/w PVDF in NMP solution into the grinded sample. A dry 1cm X 1cm
carbon paper was weighed and the slurry was applied evenly on the surface of the weighed carbon paper and dried
under vacuum in a vacuum oven at 400C overnight to remove the NMP solvent and trapped air. The dried carbon
paper was weighed to obtain the mass of sample applied onto it. It was then set up as a working electrode, along
with a platinum wire counter electrode, and a silver chloride reference electrode. CV was then conducted using
Solartron Analytical Model 1470E with the following scan rates: 5, 10, 20, 40, 80, 100, 200, 500 and 1000mV.s-1
and potential window of -0.5 to 1.5V. From the obtained voltammogram, the area within the graph is calculated
using software and the specific capacitance is calculated with the equation shown in figure 2.3.1. This procedure
was repeated for each sample.
Figure 2.3.1: Equation for calculating specific capacitance
2.4 X-ray Diffraction (XRD) test Some amount of a sample was powdered with a mortar and transferred onto the centre of a specimen holder
and was subsequently pressed down using a glass slide to flatten the sample. The specimen holder was then
transferred into the specimen stage of Shimadzu LabX XRD-6000 X-ray Diffractometer and the test was carried out
using a pre-set programme.
3. Results and discussions
3.1 Morphology When the samples were viewed under SEM, we observed various morphologies which are recorded in table
3.1.1. In total, 3 morphologies of manganese oxide NPs were obtained and they are flower, hexagonal plate and
nanorods which can be illustrated in figure 3.1.2.
3.2.1 Time effect
From the morphologies obtained in our sample, it was observed that as synthesis time increases, the favoured
morphology shifts from flower to hexagonal plates followed by nanorods. Based on the observations recorded in
table 3.1.1, part of this trend can be observed from sample B1 to B6 and C1 to C6. Flower and hexagonal plate NPs
were observed at 45mins synthesis with flower morphology being most prominent. However as synthesis time
increases, only hexagonal plate morphologies were observed. Another part of this trend can be observed in samples
from C7 to C9 where both flower and hexagonal plates NPs were observed in 45mins synthesis time but at higher
synthesis time, flower morphology were no longer observed while nanorods morphologies were observed instead.
However, this trend could not be observed from samples ranging A1 to A9 which was synthesised using 0.4ml
of 0.25M KMnO4 and from samples from B7 to B9 which have common parameters of 200°C and 1ml of 0.25M
-
18
KMnO4. This could be due to the smaller amount of 0.25M KMnO4 used and the applied synthesis temperature
parameter having a more dominant effect over the time parameter thus preventing this trend from expressing.
Table 3.1.1: Morphologies of NPs observed in samples with varying synthesis time, temperature and volume of
0.25M KMnO4
3.1.2 Temperature effect
It was observed that an increasing synthesis temperature results in a similar trend where the favoured type of
morphology shifts from flower to hexagonal plates followed by nanorods. This could be observed from sample B1,
B4, B7, C1, C4 and C7. Flower morphology were prominent at lower synthesis temperatures but as synthesis
temperature increases towards 200°C, hexagonal plates morphology was observed to have higher prominence than flower. From sample C2, C5, C8, C3, C6 and C9, it was observed that at lower synthesis temperatures of 120°C and
160°C, only hexagonal plate morphology were found. However at 200°C, nanorods were observed also. This trend
was not observed in other samples. This may also probably due to the other parameters having a more dominant
effect over the temperature parameter which as a result prevented this trend from expressing.
3.1.3 Effect of volume of 0.25M KMnO4
The volume of KMnO4 used was observed to affect the morphology of NPs as well. An increase in the volume
of KMnO4 used was observed to lead to flower and nanorods morphologies. This can be observed in samples A1, B1,
C1, A4, B4, C4, A7, B7 and C7 where the morphology of the greatest prominence shifts from hexagonal plates to
flower morphology with increasing volume of KMnO4.Samples A8, B8, C8, A9, B9 and C9 on the other hand shows
that with increasing volume of KMnO4 the morphology of the greatest prominence shifts from hexagonal plates to nanorods morphology. Similarly as previously discussed parameters, the possibility of a dominant effect of other
parameters could have prevented other samples from expressing these trends. It has been noted that only at 0.4ml of
0.25M of KMnO4, some samples were observed to be absent of NPs. As KMnO4 is the precursor to MnO2, the low
-
19
amount of KMnO4 used may have result in low yield of MnO2 which could be too low for significant precipitation
of MnO2 to form NPs.
a) b)
c)
Figure 3.1.2: a) Flower morphology MnO2 NP b) Hexagonal plate morphology MnO4 NP c) Nanorods
Morphology MnO2 Nanoparticle NP
3.1.4 Discussion: Comparing to other literature such as one by Suh et al., they reported that an elevated temperature and longer
synthesis time favoured the growth of nanorods morphologies over the flower morphologies [5]. In another study
conducted by Subramanian et al. it was also found that with increasing synthesis time the flower NPs forms into
hexagonal plate NPs first before subsequently forming nanorods NPs [4]. These findings are consistent with our
results however the difference in our results is the additional presences of hexagonal plates morphology which
suggest that the morphological transformations of our NPs in most of our sample were incomplete since hexagonal
plate morphologies were found to be intermediates of the transformation from flower to nanorods morphologies. For
how morphologies are affected by concentration of reactants used, we are at the moment unable to find any similar
studies for comparison as our project may be one of the first to conduct morphology studies using this parameter.
It is known that pseudocapacitance is proportional to the surface area of NPs and surface area of flower
morphology is the highest followed by hexagonal plate morphology and subsequently nanorods. It thus can be deduced that decreasing temperature and time could lead to pseudocapacitors with higher specific capacitance as
flower morphology is favoured with these parameters.
3.2 Cyclic voltammetry test
-
20
Each sample was tested with different scan rate: 5, 10, 20, 40, 80, 100, 200, 500 and 1000mV.s-1. The specific
capacitance results obtained for each sample are shown in table 3.2.1. For this discussion, results obtained from
40mV.s-1 scan rate have been selected for discussion and comparison.
Using statistics, we identified the range in which the specific capacitance results are not considered an outlier.
Median of data: 171.25 1st Quartile: 129.44; 2nd Quartile: 219.75
Interquartile range: 219.75 – 129.44 = 90.31
219.75 + 90.31 X 1.5 = 355.215; and 129.44 – 90.31 X 1.5 = -6.025
Table 3.2.1: Sample’s specific capacitance (F/g)
Parameter 1: Parameter 2: Parameter 3: Time (Min)
Volume of
0.25M
KMnO4 (mL)
Temperature (°C) 45 120 180
0.4
120
A1
162.20 A2
128.32 A3
277.07
160
A4
207.36 A5
219.75 A6
169.57
200
A7
468.36 A8
114.56 A9
151.51
1
120
B1
212.31 B2
129.44 B3
140.01
160
B4 159.34
B5 174.17
B6 171.25
200
B7
121.94 B8
209.97 B9
114.16
2
120
C1
187.97 C2
99.23 C3
80.45
160
C4
244.39 C5
334.60 C6
275.36
200
C7
170.21 C8
323.17 C9
184.57
Thus the range of specific capacitance in which the result is acceptable is 0 to 355.215F/g. Therefore sample
A7 with 468.36 F/g is an outlier data and will be discussed separately from other samples. Based on the CV results
obtained, outlier sample A7 produced the highest specific capacitance of 468.36 F/g. However, the SEM images of
this sample as shown in figure 3.2.2 did not show any presence of NP.
This observation of achieving very high specific capacitance despite absence of NPs in SEM images may be a
result of non-uniform distribution of NPs throughout the sample as shown by the two red circles in figure 3.2.3
where one circle has NP present while the other have none. Thus for that sample, the specimen sent for CV test may have a large amount of NPs present while the specimen sent for SEM analysis may have extremely low amount NPs
found on it. Therefore more tests have to be carried out in future to investigate this anomaly.
From table 3.2.1 there are no observable trends as to how different parameters affect the specific capacitance of
the produced pseudocapacitor material. However from the results it was observed that samples with higher specific
capacitance are most frequently produced when using 120 mins synthesis time or 160°C synthesis temperature or
2ml of 0.25M of KMnO4. This could help in understanding the optimum conditions to synthesise manganese oxide
NPs with optimum specific capacitance.
-
21
Based on the results in table 3.2.4, there is no observable trend which suggests that the NP’s morphology plays
a part in the specific capacitance. The lowest and highest specific capacitances were obtained by samples with
hexagonal plate morphology. Furthermore, samples with nanorods morphology managed to exhibit higher specific
capacitance than samples with flower morphologies. This is unexpected as it is theoretically expected that flower
morphologies with higher surface area would exhibit higher specific capacitance. Furthermore, many journals during
literature review also indicated that flower morphologies would exhibit higher specific capacitance.
Figure 3.2.2: SEM images of outlier sample
Figure 3.2.3 SEM image showing non uniform distribution
Table 3.2.4: Range of specific capacitance achieved by various morphologies
Morphology Range of specific capacitance (F/g)
Flower 170.21 – 212.31
Hexagonal plates 80.45 – 334.60
Nanorods 184.57 – 323.17
A possibility as to why hexagonal plates and nanorods morphology achieved better specific capacitance than
flower morphology despite theory and journals supporting flower as the morphology that would achieve superior
-
22
specific capacitance; may be due to the fact that embedment and amount of NPs produced playing a significant role
in determining the resulting specific capacitance.
3.3 X-ray diffraction test
a)
b) c) Figure 3.3.1: a) XRD graph of the various morphologies in our samples b) XRD graphs of birnessite MnO2; c)
XRD graphs of nsutite MnO2
By comparing the flower morphology XRD data with the birnessite crystal structure XRD data recorded by
Zhu et al. in figure 3.3.1, it is found that MnO2 in NPs with flower morphology are arranged in birnessite crystal
structure as they have common peaks at 25°, 37° and 66° [7]. On the other hand, by comparing the hexagonal plates
and nanorods morphology XRD data with the nsutite crystal structure XRD data recorded by Julien and Massot in
figure 3.3.1, it is found that MnO2 in NPs with the hexagonal plates and nanorods morphology are arranged in
birnessite crystal structure as they have common peaks at ~25°, 37°, 43° and 56° [2].
Birnessite, in theory and from various literature exhibits superior capacitance over nsutite crystal structure as the crystal structure of birnessite has better ion-exchange, adsorption, intercalation and higher surface area available
on the crystal. But as discussed in cyclic voltammetry section under effects of NP’s morphology on specific
capacitance, it was found that flower morphology achieved the lowest specific capacitance compared to nanorods
and hexagonal plate morphologies. This may be due to other possible factors such as amount of NP present and
degree of embedment which may have a significant or dominant effect over that of crystal structure.
-
23
4. Conclusions
In our search of optimising the specific capacitance of MnO2-CNT nanocomposite, success was made in
producing MnO2 nanoparticles with flower, hexagonal plates and nanorods morphologies for our studies. Through
XRD, flower morphology was found to possess birnessite crystal structure while hexagonal plates and nanorods
morphologies were found to possess nsutite crystal structure. By comparing results between the different parameters used and the morphologies observed, it was observed that as synthesis temperature and time increases the favoured
morphology shifts from flower to hexagonal plates to nanorods while with increasing volume of KMnO4 used, it
was observed to lead to flower and nanorods morphologies. It was expected that samples with flower morphology
NPs would achieve higher specific capacitance as it possess a higher surface area and a birnessite crystal structure
but results from CV tests show otherwise.
It was therefore hypothesised that there may be other factors such as degree of NPs embedment and amount of
NPs synthesised that may have a more dominant influence on the specific capacitance of the MnO2-CNT
nanocomposite and thus further investigation into other possible factors should be carried it out. Last but not least a
discovery that may potentially make progress in our goal to optimise the specific capacitance of MnO2-CNT
nanocomposite, it was observed that higher specific capacitance was frequently shown to be achieved at 1600C or
120 minutes or using 2mL of 0.25M potassium permanganate where the highest specific capacitance obtained was
335F/g.
5. References
[1] Cottineau T., Toupin M., Delahaye T., Brousse T., Belanger D.: Nanostructured transition metal oxides for aqueous hybrid
electrochemical supercapacitors. Applied Physics A, 82, 599-606 (2006).
[2] Julien C. M., Massot M.: Vibrational Spectroscopy of Electrode Materials for Rechargeable Lithium Batteries III. Oxide Frameworks. Proceedings of the International Workshop "Advanced Techniques for Energy Sources Investigation and Testing", (2004).
[3] Raymundo-Piñero E., Khomenko V., Frackowiak E., Beguin F.: Performance of Manganese Oxide/CNTs Composites as Electrode Materials for Electrochemical Capacitors. The Electrochemical Society, 152, A229-A235 (2005).
[4] Subramanian V., Zhu H., Vajtai R., Ajayan P. M., Wei B.: Hydrothermal Synthesis and Pseudocapacitance Properties of MnO2 Nanostructures. Journal of Physical Chemistry, 109, 20207-20214 (2005).
[5] Suh C. P., Suk F. C., Chian Y. L.: Controlled Synthesis of Manganese Dioxide Nanostructures via a Facile Hydrothermal Route. Journal of Nanomaterials, 2012, (2012).
[6] Wei W., Cui X., Chen W., Ivey D. G.: Manganese oxide-based materials as electrochemical supercapacitor electrodes.
Chemistry Society Review, 40, 1697-1721 (2011). [7] Zhu J., Shi W., Ni X., Rui X. Tan H., Lu X., Huey H. H., Ma J., Yan Q.: Oxidation-Etching Preparation of MnO2 Tubular
Nanostructures for High-Performance Supercapacitors. American Chemical Society Applied Materials Interfaces, 4, 2769-2774 (2012).
Mr. Ho Thiam Aik is currently working in the Singapore Polytechnic, Department of Student Development and
Alumni Relations as Alumni Manager. Prior that, he has 17 years of teaching experience in the School of Chemical
and Life Sciences. His basic degree was in Materials Engineering from Nanyang Technological University, and he
obtained his MSc in Environmental Safety and Health from National University of Singapore. His areas of
specialization include Corrosion Science, Analytical Chemistry and Workplace Safety and Health.
-
24
Synthesis and Characterisation of Anode Nanomaterials for
Lithium Ion Batteries
H. T. Tan1, Q. Y. Yan Alex1, Y. Y. Kee2, Z. W. Chan2, C. S. Leow Kevin2 and T. A. Ho2
1School of Material Science and Engineering Nanyang Technological University
2Diploma in Materials Science
School of Chemical & Life Sciences
Singapore Polytechnic
Abstract Iron (III) oxide nanoparticles are prospective alternative anode materials for lithium ion batteries owing to their higher theoretical capacity, excellent corrosion resistance and eco-friendly nature. Hydrothermal synthesis using solution-based reactions will be applied to produce well-controlled size,
morphology and composition iron oxide nanoparticles. In order to discover the optimum conditions in which the best electrochemical performance can be obtained, the nanoparticle morphology, size, crystal structure, and corrosion rate under varying parameters of synthesis time, temperature as well as the concentration of iron source were analysed and studied respectively using scanning electron microscopy, X-ray diffraction, cyclic voltammetry and electrochemical impedance spectroscopy. According to the experimental results, it
was found that the sample with optimum parameter was synthesised at 120°C for 4 hours using 1 mmol of iron (III) chloride.
Keywords: Iron (III) oxide nanoparticles; theoretical capacity; corrosion rate; scanning electron microscopy.
1. Introduction
So far, the vast need for long-lasting high capacity batteries still could not be satisfied by current energy storage technology. Studies on nano-structured materials with different morphologies are of great interest in the field
of lithium ion batteries because of their higher lithium-ion diffusion coefficients and larger contact area between the
electrode and electrolyte [1]. Among all possible anode materials, iron (III) oxide, Fe2O3 stands up as one of the
safer, abundant, less expensive and higher capacity alternatives. The Fe2O3 crystal lattice can store six Li ions per
formula unit and its theoretical capacity is as high as 1005 mAh/g, versus 370 mAh/g of the conventional graphite
anode [3]. The reversible electrochemical reaction of Fe2O3:
Fe2O3 + 6Li ⇌ 3 Li2O + 2Fe
In this experiment, iron (III) chloride hexahydrate, FeCl3.6H2O reacts with water and oxygen to form iron (III)
oxide nanoparticles through hydrothermal synthesis. In the presence of strong acidic solution, hydrochloric acid (HCl), the iron precursors are being hydrolysed and oxidised its oxide derivatives. Due to the hydrated nature of iron
(III) chloride, the iron (III) oxide formed is hydrated as well [2].
4 Fe + 3 O2 + 2 H2O → 2 α-Fe2O3.H2O / 4 α-FeO(OH)
Hydrated iron (III) oxide or iron (III) oxyhydroxide is also written as α-FeO(OH). The water content within in
the crystal structure of iron (III) oxyhydroxide can be eliminated through dehydration. The most common way is to
perform thermal processing called annealing. The temperature required for dehydration to take place above 200°C
and it is indicated that a-FeO(OH) precursors can be completely transformed into hematite at 300°C [3].
2 α-FeO(OH) → Fe2O3 + H2O
When iron (III) ions, Fe3+ first reacts with water molecules in solution, six-line ferrihydrite (Fe5HO8.4H2O)
nanodots are formed. After that, ferrihydrite (Fe5HO8.4H2O) nanodots are converted to goethite, α-FeO(OH)
-
25
nanodots through the hydrolysis process as the hydrothermal reaction goes on. Goethite nanodots aggregated and
grew into nanorods through “Ostwald ripening” mechanism [4]. Through annealing, goethite nanorods were sintered
at 300°C for 2 hours in air and transformed to mesoporous hematite α-Fe2O3 nanorods. The formation of mesopores is due to removal of hydroxide, OH groups when FeO(OH) was transformed to α-Fe2O3. These α-Fe2O3 nanorods are
further transformed into nanospheres which are more stable via “Ostwald ripening” process [4].
2. Experimental
2.1 Preparation of Anode Material via Hydrothermal Synthesis
Figure 1: Procedure of Hydrothermal Synthesis
Table 1: Samples and their respective parameters
Parameter 1: Synthesis Time (Hour)
4 5 6 12
Par
ame
ter
3:
Co
nce
ntr
atio
n o
f Ir
on
(II
I) C
hlo
rid
e (
mm
ol)
0.5
Par
ame
ter
2:
Syn
the
sis
Tem
pe
ratu
re (
°C) 1
00
- - - -
12
0
120-III-0.5-4 120-III-0.5-5 120-III-0.5-6 120-III-0.5-12
15
0
150-III-0.5-4 150-III-0.5-5 150-III-0.5-6 150-III-0.5-12
1
Par
ame
ter
2:
Syn
the
sis
Tem
pe
ratu
re (
°C) 10
0
100-III-1-4 100-III-1-5 100-III-1-6 100-III-1-12
12
0
120-III-1-4 120-III-1-5 120-III-1-6 120-III-1-12
15
0
150-III-1-4 150-III-1-5 150-III-1-6 150-III-1-12
1.5
Par
ame
ter
2:
Syn
the
sis
Tem
pe
ratu
re (
°C) 10
0
- - - -
12
0
120-III-1.5-4 120-III-1.5-5 120-III-1.5-6 120-III-1.5-12
15
0
- - - -
5) 0.56g Sodium Nitrate
6) 30mL Distilled Water
7) 80 µL Hydrochloric Acid
8) Iron (III) Chloride
9) Carbon Felt
Autoclav
e
Oven Cooling down
Cleaning and drying
of carbon felt
-
26
Desired concentration of variable iron (III) chloride and 0.56g (1 mmole) of sodium nitrate were dissolved in
30 mL of distilled water inside an autoclave’s Teflon capsule followed by addition of 80µL of hydrochloric acid.
After several minutes of ultrasonic dispersing, a carbon felt, which acts as current collector was fully dipped into the
solution. The autoclave with its Teflon capsule inside was then subsequently placed and heated in an oven at the
desired temperature and time. After cooling down for 2 hours, the carbon felt was taken out from the solution,
washed carefully with distilled water, and eventually dried at 70°C overnight in the oven. This procedure as illustrated in Figure 1 is repeated for all the 28 samples with their corresponding parameter.
2.2 Annealing The as-prepared sample was calcined in a quartz tube at 400°C for 2 hours with a heating rate of 10°C min-1 in
Ar atmosphere.
2.3 Sample Characterizations
JOEL JSM-7600F scanning electron microscope was employed to examine the morphology of the sample.
The crystal structure phases of the samples were characterised by Shimadzu X-ray diffraction (Cu-Kα, λ = 1.5406 Å)
from 5° to 80° at a step size of 0.02 s-1.
2.3 Electrochemical Measurements To test the anode performance of synthesised materials, CR 2025 coin cells were made using Celgard 2400 as
the separator and the electrolyte was 1 M LIPF6 in 1:1 mixture of ethylene carbonate and diethyl carbonate. The coin
cells were assembled inside an argon-filled glove box with oxygen and water contents below 1 and 0.1 ppm,
respectively. Li-metal was used as the counter and reference electrode. The working electrode was fabricated by the
active material (iron oxide) on the carbon felt. Galvanostatic charging and discharging tests were conducted using a
battery tester (1470E Eight Channel Potentiostat/Galvanostat) at different current densities at room temperature.
Cyclic voltammetry was performed using an electrochemical workstation (CHI 660C) from 1 mV to 3 V at a
scanning rate of 0.2 mV s-1. Electrochemical impedance spectroscopy was carried out at the scan mode of 10 mV s-1.
3. Results and Discussions 3.1 Scanning Electron Microscopy
(a) (b)
(c)
Figure 2: (a) *Nanodots morphology NPs (b) Nanorods morphology NPs; (c) Nanospheres morphology NPs
-
27
Table 2: Morphologies of NPs observed in samples with varying synthesis time, temperature and
concentration of iron (III) chloride
Samples NPs’ Morphology
100-III-1-4 Nanorods & Nanodots
100-III-1-5 Nanospheres
100-III-1-6 Nanospheres
100-III-1-12 Nanospheres
120-III-0.5-3 Nanorods & Nanodots
120-III-0.5-4 Nanorods & Nanodots
120-III-0.5-5 Nanospheres
120-III-0.5-6 Nanospheres
120-III-0.5-12 Nanospheres
120-III-1-3 Nanorods & Nanodots
120-III-1-4 Nanorods
120-III-1-5 Nanorods, Nanodots & Nanospheres
120-III-1-6 Nanospheres
120-III-1-12 Nanospheres
120-III-1.5-3 Nanorods & Nanodots
120-III-1.5-4 Nanorods
120-III-1.5-5 Nanorods, Nanodots & Nanospheres
120-III-1.5-6 Nanorods &Nanospheres
120-III-1.5-12 Nanospheres
150-III-0.5-4 Nanorods & Nanodots
150-III-0.5-5 Nanospheres
150-III-0.5-6 Nanospheres
150-III-0.5-12 Nanospheres
150-III-1-3 Nanorods & Nanodots
150-III-1-4 Nanorods & Nanodots
150-III-1-5 Nanorods, Nanodots & Nanospheres
150-III-1-6 Nanorods &Nanospheres
150-III-1-12 Nanospheres
*For simplicity purpose, we classified them as “nanodots”. To be more precise, they should be termed as nanorods with shorter length, or nanorods with lower aspect ratio.
3.1.1 Morphology
-
28
Through SEM analysis, iron oxide nanoparticles (NPs) of three morphologies were being produced. These
morphologies include *nanodots, nanorods and nanospheres which are illustrated in Figure 2 and summarised in
Table 2.
3.1.2 Size/Dimension
Table 3: Dimension range of NPs observed in samples with varying synthesis time, temperature and concentration of iron (III) chloride
Samples NPs’ Dimension Range (nm)
100-III-1-4 Small & Medium-sized NPs
100-III-1-5 Small & Medium-sized NPs
100-III-1-6 Medium-sized & Large NPs
100-III-1-12 Medium-sized & Large NPs
120-III-0.5-3 Small NPs
120-III-0.5-4 Small NPs
120-III-0.5-5 Small NPs
120-III-0.5-6 Small & Medium-sized NPs
120-III-0.5-12 Medium-sized NPs
120-III-1-3 Small NPs
120-III-1-4 Small NPs
120-III-1-5 Small & Medium-sized NPs
120-III-1-6 Medium-sized & Large NPs
120-III-1-12 Medium-sized & Large NPs
120-III-1.5-3 Small NPs
120-III-1.5-4 Small & Medium-sized NPs
120-III-1.5-5 Medium-sized NPs
120-III-1.5-6 Medium-sized & Large NPs
120-III-1.5-12 Large NPs
150-III-0.5-4 Small NPs
150-III-0.5-5 Small & Medium-sized NPs
150-III-0.5-6 Medium-sized & Large NPs
150-III-0.5-12 Large NPs
150-III-1-3 Small NPs
Samples NPs’ Dimension Range (nm)
150-III-1-4 Small & Medium-sized NPs
150-III-1-5 Medium-sized & Large NPs
150-III-1-6 Medium-sized & Large NPs
150-III-1-12 Large NPs
-
29
The size of iron oxide NPs formed was categorized into three ranges: small NPs (400nm). The estimated size range of NPs was determined by measuring the largest as
well as the smallest NPs in an SEM image in the aids of dimension scale shown at the bottom of image. For
nanorods, the length was taken into account. On the other hand, nanodots and nanospheres were measured in the
aspect of diameter. Refer to Table 3 below.
In conclusion, the iron oxide NPs with nanorod morphology and within a size range of 200nm or less are preferable in terms of electrochemical performance. These NPs possess the largest possible surface area for
simultaneous lithium ion insertion provided the volume of NPs is the same. Large surface area of NPs permits:
i. A high contact area with the electrolyte and hence a high lithium-ion flux across the interface. ii. Shorter distance for lithium-ion transport within the particles.
Due to these reasons, anode with smaller NPs tends to conduct a higher capacity when compared to NPs
with NPs of larger dimension.
3.1.3. Parameters’ Effects on NPs
3.1.3.1 Synthesis Time
From the SEM images, it was observed that a shorter synthesis time results in the formation of rod-like and
polygonal NPs. As the synthesis time increases, NPs formed are typically found in spherical structure. The trend where a shorter synthesis time led to nanorods and polygonal NPs can be observed from the samples which were
produced at a synthesis time of 4 hours or less. The presence of nanorods and irregular NPs was clearly visible in
SEM images of these 10 samples while spherical NPs were not noticeable.
Spherical NPs started to appear on the samples when the synthesis time increased to 5 hours. In contrast,
polygonal NPs were hardly found or totally vanished in these samples. As the synthesis time increased from this
point to 6 hours, the amount of spherical NPs increased and nanorods were getting lesser. When the synthesis time
was raised to 12 hours, all of the samples were entirely composed of spherical NPs. It can be seen that a short
synthesis time led to the formation of favourable nanorod morphology.
Generally, the dimension of NPs formed increased proportionally with the increase in synthesis time. This
trend can be well observed from parameters set: 150-III-1-3, 150-III-1-4, 150-III-1-5, 150-III-1-6 and 150-III-1-12.
At constant temperature, types and concentration of iron chloride, the NPs were growing larger when the synthesis time increased. On the sample 150-III-1-3, NPs were formed in a range of 50-150 nm. In contrast, the NPs on the
sample 150-III-1-12 were oversized in terms of nanoscale, whereby the range was 500-1300 nm. It can be seen that
decreased synthesis time caused smaller NPs of desirable scale to be formed.
3.1.3.2 Synthesis Temperature
According to SEM images, there was no observable trend stating at different temperature, NPs of different
morphologies were synthesized. An increasing temperature led to a higher retention of nanorods and polygonal NPs
after longer synthesis time. This was deduced by comparing the sample products of 1mmol iron chloride at different
temperatures. After processing at temperature of 100°C, nanorods can only be found on the sample experiencing 4
hours synthesis time. The rest was merely consisted of spherical NPs. When the processing temperature increased to
120°C, nanorods were observed on 4-hour and 5-hour samples. This structure of NPs existed on 4-hour, 5hour and
6-hour sample when the temperature further rose to 150°C. it was found that an increase in temperature resulted in higher nanorods formation possibility.
It was inferred that the dimension growth of NPs was accelerated when the temperature rose. This trend can be
observed from the change in NPs size of 0.5mmol iron chloride samples. At 120°C, the NPs formed on the anode
samples were either considered small or medium-sized, with dimension up to 400nm. Only small NPs were
observed on 3-hour, 4-hour and 5-hour samples and no large NPs exceeding 400nm was found on 12-hour sample.
On the other hand, medium-sized NPs were observed from 5-hour sample when it went through a processing
temperature of 150°C. Furthermore, large NPs which have size of 400nm and above started to be formed on 6-hour
sample. However, this trend was not observed when other parameters were used. This may due to more dominating
effect of synthesis time and iron chloride concentration over the temperature parameter which as a result prevented
this trend from being significantly observed. It was found that dimension of NPs decreased at lower synthesis
temperature.
3.1.4 Concentration of Iron (III) Chloride
-
30
3.1.4.1 Concentration effect on Morphology
There was no trend showing NPs of different morphologies were produced when concentration of iron chloride
varied. According to the SEM images, it was discovered that at higher FeCl3 concentration, the nanorod NPs stood a
greater chance to be fabricated. This was clearly showed by comparing the morphology of NPs formed on the basis
of varying FeCl3 concentration. When 0.5mmol of FeCl3 was used, nanorod NPs were only presented on the 3 hours and 4 hours samples. In contrast, upon the addition of 1.5mmol of FeCl3, nanorod NPs were found on the samples
with synthesis time of 3 to 6 hours. It can be seen that nanorod NPs were more likely to be produced at a higher iron
(III) chloride concentration.
3.1.4.2 Concentration effect on Dimension Similarly, iron chloride concentration did not alter the size of NPs drastically. Nevertheless, it was observed
that NPs enlarged faster when iron chloride concentration increased. When 0.5mmol of iron chloride was added to
undergo 3 and 4 hours synthesis process, the NPs formed on the anode possessed dimension of lower than 200nm.
On the other hand, medium-sized NPs where the dimension range is 200-400nm started to be formed on 4-hour-
samples when iron chloride concentration was 1.0mmol and 1.5mmol. It was found that NPs were growing smaller
at a faster rate when iron (III) chloride concentration decreased.
Table 4: Variables requirement to obtain NPs of desirable morphology and size
Variables Desirable Morphology
(Nanorods)
Desirable Size
(Small:
-
31
Figure 3: SEM image of the annealed sample
3.2 X-ray Diffraction (XRD) Analysis
(a) (b)
Figure 4: (a) XRD pattern of sample 120-III-1-4 (b) “Matching” with XRD analysis software
(a) (b)
Figure 5: (a) XRD pattern of sample 120-III-1-4 after annealing (b) XRD pattern of commercial
hematite sample
3.3 Cyclic Voltammetry (CV) Test
(a) (b)
Figure 6: (a) Cyclic voltammogram curves of sample 120-III-1-4 (b) Cyclic voltammogram curves of
sample 120-III-1-4 after annealing
-
32
Figure 6(a) shows the CV curves of sample 120-lll-1-4 in the first three cycles in the voltage range from 3 to
0.001 V at a scan rate of 0.2 mV s-1. From the CV curves, cathodic peaks located at 2.1 V and 1.58 V could be
attributed to the intercalation of lithium before the reduction of iron oxide NPs. These two peaks disappeared in the
subsequently cycles, indicating their irreversibility in nature. In addition, a peak located at 0.83 V represent the
conversion of Fe3+ to Fe2+ and continue reduction of Fe2+ to Fe0. In the anodic scan, the oxidation peaks at 1.1 and 1.58 V correspond to the oxidation of Fe0 to Fe2+ and Fe2+ to Fe2O3, respectively. The presented peaks match the
theory of reduction and oxidation peaks of the iron oxide versus lithium reported in the literature. However, in the
following second and third cycles, the cathodic peak and anodic peaks with reduced peaks intensity become
indistinguishable. The reason is the certain irreversibility of the redox reactions during charging and discharging.
Figure 6(b) demonstrates the CV curves of the annealed sample 120-lll-1-4. It shows similar voltage locations of
cathodic and anodic peaks with sample 120-lll-1-4 because the annealing process only removes the crystalline water
from the iron-oxide NPs but does not change its phase. Hence, the cathodic and anodic peaks are remaining at the
same voltage point.
(a) (b)
Figure 7: (a) Cycling performance chart of sample 120-III-1-4 (b) Cycling performance chart of sample
120-III-1-4 after annealing
The cycling performance chart 7(a) is obtained from the sample of 120-lll-1-4 at 1 C. The sample show initial
discharge specific capacities of 869, 813 and 805 mAh/g, and decreases to 754mAh/g at the 55th cycle at a low current density of 1 C. After 55 cycles, the capacity of the coin cell is still able to maintain 87% of its original
capacity. It is considered as good and stable performance.
Under low current density charging and discharging, the cyclic performances show a good stability up to 60
cycles which maintain the specific capacity around 800mAh/g. The SEM results revealed that the nanorods structure
of the iron oxide NPs was the main reason that caused the excellent stability of the coin cell. The underlying reason
is because of the high surface area of the active materials for lithium ions intercalation. On the other hand, long
reaction time tends to cause the hydrothermal synthesis form bigger particles in sphere structure which can be
observed from SEM results. The formations of large sphere particles greatly affect the capacity and stability
performance of the coin cell. When the current rate is increased from 1 C to 5 C with 1 C interval value, there is a decreasing value of
discharge-charge specific capacities as show in the figures above. There were 11 charging-discharging cycles at
each different current density for the coin cell testing. Hence, there are a total of at least 55 cycles for the sample.
The higher current density used for measurement will decrease the specific capacity of the cell because the rate of
conversion reaction (iron oxide NPs) occurs at the electrode is not able to keep pace with the charge/discharge rate.
In this scenario, the capacity contributed from the redox reaction of the active materials becomes the determining
factor that limits the specific capacity of a coin cell.
The sample exhibits excellent cycling stability because of nano-dimension of hydrated Fe2O3 nanorods that
give rises to specific surface area. In order to obtain better rate performance, annealing process was conducted to
-
33
remove all the water molecules within the crystal lattice without sacrificing their uniformity and initial morphology.
Figure 7(b) shows 55 cycles cycling performance chart of annealed 120-lll-1-4 sample. After annealing, the sample
was carried out CV test with the same parameters and conditions. The discharge specific capacities of the samples in
the first 3 cycles are 1183, 1111, and 1093 mAh/g. At the 55th cycle, the annealed sample gives a discharge specific
capacity of 1010 mAh/g which retains 85% of its original specific capacity.
From the preliminary results and analyses, the annealed product gives an improved performance in term of
higher specific capacity and good charging-discharging coin cell stability.
3.4 Electrochemical Impedance Spectroscopy (EIS)
(a) (b)
Figure 8: (a) EIS result of sample 120-III-1-4 (b) EIS result of conventional coin cell
Due to time constraint, we only managed to investigate the corrosion rate of battery produced from the
optimum formulation (120-lll-1-4). The annealed sample is still undergoing CV test which required longer testing
time. However, we believed that annealed sample should be able to give greater corrosion resistance as the water
molecules in the crystal lattice can be removed completely by means of annealing. Conventional graphite anode material plays the role of reference during this EIS test for comparison purpose. Based on the EIS results, the
optimised battery has an average corrosion rate of 7 mm/yr and conventional lithium ion battery shows a higher
average corrosion rate of 10 mm/yr. These corrosion rate results may seem high because we are simulating the
worst case scenario whereby the whole battery is completely immersed in tap water
a : atomic weight
icoor : corrosion current density, µA/mm2
n : number of electrons lost (valency change)
D : density, g/cm3
K : a constant depending on the unit of corrosion rate
Unit of Corrosion Rate K Value
mpy 0.129
µm/yr 3.27
Mm/yr 0.00327
The values of parameters are as below:
-
34
K= 0.00327 mm/yr D= 2.7 g/cm3 n= 3
a= 27 g/mol Area of coin cell= 314.16 mm2
Figure 9: Equation for calculating corrosion rate
Current density of samples, µA:
1 2
Hydrated Fe2O3 NPs 200 µA 180 µA
Graphite 300 µA 250 µA
icorr of samples:
1 2
Hydrated Fe2O3 NPs 0.637 µA/mm2 0.573 µA/mm2
Graphite 0.955 µA/mm2 0.796 µA/mm2
Corrosion rate:
1 2 Average
Hydrated Fe2O3 NPs 6.94 mm/yr 6.25 mm/yr 7 mm/yr
Graphite 10.41 mm/yr 8.68 mm/yr 10 mm/yr
The main reason of hydrated Fe2O3 NPs has a lower corrosion rate is due to the formation of protective
passivation layer. The layer protects the anode from reacting with the surrounding oxygen. On the other hand,
graphite anode is reactive and it tends to form a weak, unstable porous layer, known as solid electrolyte interphase
(SEI). During charging and discharging, the SEI will partially dissolve into the electrolyte which causes the
corrosion of anode and reduction of insertion places for electron ions. Hence, the corrosion rate of LIB using
graphite anode is greater.
4. Conclusion
In our approach to synthesise a ‘greener’ anode material for battery, it was observed that the optimised sample
was made by using 120°C, Fe (III) as iron source, 1 mmol of iron source, and 4 hours. The optimised anode sample
demonstrated the following properties:
i. The sample, 120-lll-1-4 has a hydrated hematite crystal structure, (Fe2O3•H2O). ii. The formation of hydrated hematite crystal structure, (Fe2O3•H2O) is arranged in nanorods morphology
which can perform excellent cycle and rate performance as the long nanorods could ensure many fast and
convenience electron transport pathway, thus enhancing the electronic conductivity leading to improved
electrochemical performance of higher specific capacity and good charging-discharging stability. The in-
situ growth of Fe2O3•H2O on carbon felt could ensure fast electron transport pathway between the active
materials and the current collector, leading to improved electrochemical performance. iii. The corrosion rate of Fe2O3 NPs is lower as compared to graphite anode material used in conventional LIB.
Furthermore, it was found that the water content in the crystal lattice can be removed completely by means of
annealing at 400°C for 2 hours under argon atmosphere. This demonstrates the probability of transforming the
hydrated form of Fe2O3 to its pure phase. The annealed sample improves crystallinity of iron oxide nanoparticles
which leads to enhance the overall electrochemical performance of battery in term of rate performance and capacity.
5. References
[1] Bruce, P. G., Scrosati, B. and Tarascon, J. 2008. Nanomaterials for Rechargeable Lithium Batte